Deficiency of ATP13A3, a polyamine transporter, rescues synthetic lethality of FANCD2 ^−/− ALDH9A1 ^−/− cells. (A) Schematic summary of the metabolism-focused CRISPR KO suppressor screen. FANCD2^−/− and FANCD2^−/−ALDH9A1^−/− Jurkat cells were infected with lentivirus-containing metabolism-focused sgRNA library in biological triplicate at MOI of 0.3. sgRNA representation of 500x was maintained throughout the experiment. sgRNAs after 14 population doublings were assessed to determine genetic perturbations that rescue synthetic lethality of FANCD2^−/−ALDH9A1^−/− Jurkat cells. (B) FANCD2, ALDH9A1, γ-H2AX, H2AX, and cleaved caspase 3 western blot of Jurkat clones used in the experiments. Vinculin was used as a loading control for each membrane. (C) cDNA-expressing WT ALDH9A1 and YFP was overexpressed in WT, ALDH9A1^−/−, FANCD2^−/−, and FANCD2^−/−ALDH9A1^−/− cells, and their relative growth was compared with luciferase-GFP–overexpressed counterparts by monitoring YFP/GFP ratio using flow cytometry. One experiment in biological triplicate was performed. One-way ANOVA followed by Dunnett’s multiple comparison test was performed (ns, not significant; *, P < 0.05; ***, P < 0.0005). Error bars, SD. N = 3. (D) A volcano plot showing log₂ fold change (log₂FC) value and P adj value between FANCD2^−/− and FANCD2^−/−ALDH9A1^−/− at the sgRNA level. Only sgRNAs with P < 0.01 were used. Ten out of ten sgATP13A3s were significantly enriched in FANCD2^−/−ALDH9A1^−/− condition and marked as indicated. Values for all data points in this graph are in Table S5. (E) A competition assay using FANCD2^−/− and two independent FANCD2^−/−ALDH9A1^−/− clones. The ratio of mCherry/GFP normalized to baseline was progressively increasing in FANCD2^−/−ALDH9A1^−/− clones with mCherry-sgATP13A3 compared with FANCD2^−/−ALDH9A1^−/− clones with mCherry-sgCTRL. There were no differences between mCherry-sgsgATP13A3 and mCherry-sgCTRL in FANCD2^−/− cells. Two independent experiments were performed. A representative figure is shown. Error bars, SD. N = 3. (F) A competition assay, as shown in Fig. 5 E, with ALDH9A1^−/− cells included. Growth advantage was not observed in ALDH9A1^−/− cells targeted by sgATP13A3. One experiment in biological triplicate was performed. Error bars, SD. N = 3. (G) Cumulative population doublings of FANCD2^−/−ALDH9A1^−/− clones (dKO) as well as FANCD2^−/−ALDH9A1^−/−ATP13A3^−/− clones (tKO), and their WT and single KO controls. Triple KO cells had improved growth rates. Two independent experiments were performed. Statistics were performed on pooled data. Error bars, SD. N = 6. (E–G) One-way ANOVA followed by multiple comparison test was performed for the last time point. (***, P = 0.0003; ****, P < 0.0001). (H) FANCD2, ALDH9A1, ATP13A3, γ-H2AX, H2AX, and α-tubulin western blot of Jurkat clones used in the experiments. ATP13A3 bulk KO was performed by Cas9-sgRNA-RNP nucleofection of FANCD2^−/−ALDH9A1^−/− cells. ATP13A3 indel percentage from the bulk cells assessed by Sanger sequencing and analysis by Synthego ICE were 89% and 91% for dKO c#2–5 and dKO c#2–18, respectively. dKO, double KO. Source data are available for this figure: SourceData F5.